Many solid and liquid substances naturally emit fluorescent radiation when irradiated with ultraviolet light. However, the radiation may fall within wide wavelength bands of low intensity. In the case of many natural objects, observations are partially obscured by natural fluorescence emanating simultaneously from many different compounds present in the sample under examination. In imaging devices such as microscopes, therefore, it is known to employ a filter for a selected UV wavelength band to screen out undesired fluorescence emanating from the object under observation.
In medical applications, a similar difficulty arises because both tumors and healthy tissue fluoresce naturally, albeit at different wavelengths. Consequently, when UV-activated fluorescence is used to detect tumors against a background of healthy tissue, identification of tumors is difficult. However, unlike most other cells of the body, tumor cells may possess a natural ability to concentrate and retain hematoporphyrin derivative dyes. Based upon this discovery, a technique was developed wherein a hematoporphyrin derivative fluorescent dye is administered and allowed to concentrate in a tumor to be examined to increase the fluorescence from the tumor as compared with that of healthy background tissue. Hematoporphyrin dyes fluoresce within a fluorescence spectrum between 610 and 700 nm, a spectrum easy to detect. However, the natural fluorescence from healthy in cells is still much more intense than that from the dyes, and has a broader fluorescence spectrum. Thus, the use of fluorescent dyes in diagnosis of tumors has not been wholly successful.
In endoscopic systems, it is also known to irradiate an internal organ with visible radiation to obtain a visible image and then to apply to the internal organ a fluorescent dye that concentrates in tumors over a period of time. The dye is allowed to concentrate, and then the internal organ is irradiated with excitation radiation for the dye to obtain a second fluorescent image. A body part having abnormal or diseased tissue, such as a cancer, may be identified by comparing an image produced by visible radiation of the internal organ with the image produced by fluorescence. To aid in visualizing the images received, endoscopic systems commonly utilize a television camera attached to a fiber optic scope having an optical guide fiber for guiding a beam from an external radiation source to the internal organ, and another optical guide fiber for transmitting a fluorescent image of the affected area to a television monitor for viewing. These two approaches are combined in a method of the type disclosed in U.S. Pat. No. 4,821,117, wherein a fluorescent dye is applied to an object to be inspected, allowed to concentrate in the tumor, and the affected site is then alternately irradiated with visible light and with radiation at the excitation wavelength of the fluorophore. Images of the object obtained independently by visible and fluorescent light using a TV camera are stored in memory, and are simultaneously displayed in a television monitor to visually distinguish the affected area of the body part from the healthy background tissue.
In another type of procedure, such as is described in U.S. Pat. No. 4,786,813, a beam-splitting system splits the fluorescence radiation passing though the optical system into at least three parts, each of which forms a respective image of the object corresponding to each of the wavelength regions received. A detector produces a cumulative weighted signal for each image point corresponding to a single point on the object. From the weighted signal values of the various points on the object, an image of the object having improved contrast is produced. This technique is used to aid in distinguishing the fluorescence from the affected tissue from that produced by normal tissue.
A still more complex method of visualizing images from an endoscopic device uses television scanning apparatus. For example, U.S. Pat. No. 4,719,508 discloses a method utilizing an endoscopic photographing apparatus wherein the endoscope includes an image sensor for successively generating image signals fed to a first frame memory for storing the image signals and a second frame memory for interlacing and storing image signals read successively from the first frame memory. The stored, interlaced image signals are delivered to a TV monitor for display to aid in visualizing the affected body part.
These prior art endoscopic systems, which rely on photographic imaging of the area of interest (i.e., via a TV monitor), while effective, have historically relied on increasingly complex and expensive equipment and substitute indirect viewing for direct viewing of the affected body part.
Certain of the fluorescent dyes that concentrate in tumors due to natural bodily processes can be excited at wavelengths corresponding to those produced by lasers to accomplish diagnostic and therapeutic purposes. Consequently, lasers have also been used in procedures utilizing endoscopic systems in conjunction with fluorescent dyes to image and treat tumors. In one embodiment of this general method, a dye is used that absorbs laser light at two different wavelengths and/or laser powers, one that excites fluorescence without generating damaging heat in the tissue, and one that generates sufficient heat in the dye to destroy surrounding tissue. U.S. Pat. No. 4,768,513, for example, discloses a procedure in which a dye is applied to a body part suspected of containing a tumor, usually by local injection. The dye is allowed to concentrate in tumors and clear from healthy tissue over a period of days, and then the body part is irradiated with alternate pulses of two light sources: a white light of a known intensity and a fluorescence-exciting laser light. To compensate for variations in intensity of the fluorescence resulting from variations in the angle of incident light, and the like, visualization of the tumor is computer-enhanced by calculating the intensity of the fluorescence with respect to the known intensity of the white light. Ablation of a tumor detected using this method is accomplished by switching the laser to the heat-generating wavelength so as to destroy the cancerous tissue into which the fluorophore has collected.
While effective for diagnosing and treating tumor, such methods have two major drawbacks. Disease states other than tumor cannot be diagnosed, and laser visualization must be delayed for a period of two days or more after administration of the fluorescent dye to allow the dye to clear from normal tissue.
Monoclonal antibodies and other ligands specific for tumors have been developed for use in diagnosis of tumors, both in tissue samples and in vivo. In addition to such ligands, certain tumor-avid moieties are disproportionately taken up (and optionally or metabolized by tumor cells). Two well-known tumor-avid compounds are deoxyglucose, which plays a telling role in glycolysis in tumor cells, and somatostatin, which binds to and/or is taken up by somatostatin receptors in tumor cells, particularly in endocrine tumors.
In such studies, deoxyglucose is used as a radio-tagged moiety, such as fluorodeoxyglucose (.sup.18 F-deoxyglucose), for detection of tumors of various types. It is believed that tumor cells experience such a mismatch between glucose consumption and glucose delivery that anaerobic glycolysis must be relied upon, thereby elevating the concentration of the radioactive tag in tumor tissue. It is also a possibility that the elevated concentration of deoxyglucose in malignant tumors may be caused by the presence of isoenzymes of hexokinase with abnormal affinities for native glucose or its analogs (A. Gjedde, Chapter 6: "Glucose Metabolism," Principles of Nuclear Medicine, 2nd Ed., W. B. Saunders Company, Philadelphia, Pa., pages 54-69). Similarly, due to the concentration of somatostatin in tumor tissue, radio-tagged somatostatin, and fragments or analogs thereof, are used in the art for non-invasive imaging of a variety of tumor types in a procedure known as somatostatin receptor scintigraphy (SRS).
Although these techniques have met with considerable success in determining the presence of tumor tissue, scintigraphic techniques are difficult to apply during a surgical procedure because of the equipment necessary for viewing the image provided by the radioisotope. Yet it is exactly at the time that the surgeon has made the incision or entered the body cavity that it would be most useful to "see" the outlines of the diseased tissue.
Thus, there is a need in the art for new and better methods that can be used to directly visualize a broad range of putative disease sites using fluorophores that are excited by UV light with or without the use of endoscopic instruments. Where visualization is by means of endoscopic devices, direct visualization (as opposed to creation of photographic images) offers the additional advantage that the equipment required is comparatively simple to use and is less expensive than the equipment required to create photographic displays from such images. In addition, there is a need in the art for a method of identifying diseased or abnormal tissue during surgical procedures so that immediate resection or biopsy of the identified tissue can be performed while the surgeon "sees" the outlines of the diseased or abnormal tissue.